Retention Economics
Part of Repeat-purchase economics
Setting a retention budget from recoverable margin
Build a provisional retention spending cap from plausible extra contribution, costs and forecast uncertainty.
Set a provisional retention budget from the extra contribution an action might create, after its costs and uncertainty.
Not all contribution earned by returning customers is available to spend.
Define the recoverable amount
Name the eligible customers, proposed action, outcome and the period in which another purchase is plausible. State the usual alternative: what service or marketing would those customers receive if this action did not run? Compare the action with that alternative, not with an imaginary absence of every other customer contact.
Estimate order contribution after expected discounts, refunds, product cost and relevant variable order costs. Record whether an incentive is already reflected in that figure; if it is, do not subtract it again as a program cost. Include costs incurred for customers who never buy, such as contact charges.
A credible prior comparison can inform the estimated difference in contribution per eligible customer. Carry over its audience, product, period and uncertainty.
Without comparable evidence, label the difference an assumption and keep spending provisional. Attributed orders and redeemed codes do not show what would have happened without the action.
Calculate a provisional cap
Amount available for fixed campaign work = eligible customers × assumed extra order contribution per eligible customer − variable program costs not already counted − uncertainty reserve.
This is a planning cap under stated assumptions, not a target spend or a promised return. A business may set a lower cap to protect cash or meet its own return requirement.
For illustration only, suppose 1,000 eligible customers are assumed to produce A$3 more order contribution each than under usual treatment. Assume the A$3 already includes any incentive on resulting orders. That gives A$3,000 of gross recoverable contribution.
After A$200 in contact costs and an A$800 reserve for forecast error, A$2,000 remains for fixed campaign work. At only A$1 of assumed extra contribution per customer, nothing remains after those same deductions. These figures are invented and come from no retailer test.
Before approval, record / Why it matters
- Incremental assumption, eligible count and period
- Sets the basis for the cap
- Costs included in order contribution
- Prevents incentive double counting
- Other contact, staff, set-up and system costs
- Shows the amount still to be funded
- Reserve, maximum spend and stop point
- Limits exposure if assumptions fail
- Comparison and review date
- Defines how the result will be assessed
Review the whole eligible group
Budget across all eligible customers, including those who never respond. Do not multiply margin per redeemed order by every recipient or add the full forecast customer value to the same expected extra order. Follow customers through a suitable buying interval; an early purchase may simply have replaced a later one.
Review the resulting difference in contribution per assigned eligible customer, including applicable program costs, returns and changes in product mix. Report the difference with its uncertainty and limit the conclusion to the tested group and period.
